Search Results

To record infrequent and elusive ECG heart arrhythmia data. Once data is recorded, patients transmit the recorded ECG data over the telephone or directly to a host PC for review.

Not Found

This document describes the acceptance criteria and study proving a device's performance. However, there is no specific study described within the provided text that directly details acceptance criteria for a new AI/algorithm, a validation study with specific performance metrics (like sensitivity, specificity, or AUC), or a comparison against human readers.

The provided text describes a 510(k) summary for a hardware modification to an ECG event recorder (ER800 Series). The "Modification Background" section states: "In connection with the modification the device performance has been validated as reported herein. Note all devices meet the same performance standards of AAMI-EC38, EN60601-1-1, EN60601-1-2, and EN60601-2-47."

This indicates that the device's performance validation is against established electrical safety and performance standards for medical devices, not against a specific set of clinical performance metrics for an AI algorithm interpreting ECGs. The primary changes are hardware-related, with a reinstatement of existing arrhythmia functionality. The comparison table focuses on hardware features and basic functionalities like storage capacity, battery life, and on-board analysis.

Therefore, many of the requested points related to AI/algorithm validation, such as sample sizes for test sets, ground truth establishment, expert qualifications, and MRMC studies, cannot be directly answered from this document.

However, based on the information provided, here's what can be extracted and inferred:

1. A table of acceptance criteria and the reported device performance:

Since the document refers to compliance with established medical device standards rather than a new algorithm's performance metrics, the "acceptance criteria" are the requirements outlined in those standards. The "reported device performance" is the statement that the device meets these standards. A direct table of numeric performance metrics is not provided for clinical outcomes like arrhythmia detection accuracy.

2. Sample size used for the test set and the data provenance:

• Sample size: Not specified. The validation is against engineering and safety standards, which typically involve testing components or the full device under various conditions, rather than a clinical "test set" of patients or ECGs.
• Data provenance: Not applicable in the context of clinical data. For testing against electrical standards, testing would be conducted in a laboratory environment.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• Not applicable as this is not an AI/algorithm validation against clinical ground truth. Validation would be performed by qualified engineers or technicians in a laboratory setting.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

• Not applicable.

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

• No, an MRMC study was not done as this is a hardware modification of an existing device, not a new AI algorithm.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• The device has "On-board analysis: Yes" which implies some automated arrhythmia detection. The text states the modification "reinstates existing arrhythmia functionality." However, a standalone performance study report for this existing (not new) functionality is not provided in this excerpt. The focus of the submission is the hardware modification's continued compliance with general device standards, not a new or revalidated performance claim for its "on-board analysis."

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc):

• Not applicable for performance relative to electrical and safety standards. If the "on-board analysis" functionality's original validation (which is pre-existing) involved ground truth, that information is not in this document.

8. The sample size for the training set:

• Not applicable as this is not a new AI algorithm.

9. How the ground truth for the training set was established:

• Not applicable.

Ask a specific question about this device